MASTER NOTES: Reader Ideas

Last week in Master Notes, I discussed Kelvin Herrera’s propensity for pitching better in saves situations than non-save situations. That topic came up on the fantastically great subscriber forums. This week’s ideas come from another of the site’s vox populi sources—the comments sections after our articles.

Reader responses put ideas into new and different perspectives. Like other BHQ analysts, I usually reply to the article comments in the article comments. But this week, I’m going to respond to the comments about the Herrera Master Notes article ... in a new Master Notes article. You might call it “a lack of imagination,” but I call it “symbiosis” or “synergy” or “a hangover.”

The first comment was:

So what can we do with this information? Nothing, it seems.

Ouch. It strikes me that the information could lead an imaginative reader to adjust how he values closers, perhaps by bumping the value of closers whose managers don’t put them into games except in save situations. The information could especially affect how the reader values Herrera in particular when making roster decisions about available closers at draft or in trade. I know I have already downgraded Herrera in my league.

There’s also a broader issue about the BHQ philosophy. I drew the possible inference here that the reader thinks I failed to say, “Here’s what you should do.” The thing is, we generally try not to say that. I don’t know anything about your context—do you have Herrera? Are you thinking of trading him or trading for him? What are the saves race and the decimals races like in your league? Etc., etc. We provide data, analysis, and comment, but we leave it to each reader to figure out how to respond for himself (or herself). To sum it up in a catchphrase: We report (objectively), you decide (rationally, based on your situation).

The next subscriber commented:

I absolutely believe in this, not only for Herrera, but for other closers...
I always dread seeing one of my closers enter in a non-save situation.

I appreciate the support, but I’m leery about the idea that “believing” in something makes it actionable. I feel like this could be an instance of confirmation bias, where a person has an established opinion and appreciates an article that confirms that position.

The issue is that the article doesn’t actually support the position in general. It does turn out that for Herrera, there appears to be something to it. But for now, it applies only to Herrera. Extrapolating from 120 innings of one closer to the entire universe of closers could be premature at best and downright misleading at worst. Based on what I found, I’d be willing at this point to place a roster bet based on Herrera’s seeming save/non-save splits, but I wouldn’t assume the same is true across the board.

I have made a note to do a research piece in the offseason to broaden the field to include all closers, and possibly over a longer period. Of course, since I wrote the note on the back of a Costco receipt, I’m not 100% confident this will ever get done. Attention Brent or Ed, if your filing system is better than mine, could you remind me about this idea in November?

The next subscriber comment:

Really enjoy your posts .... I would like a study of hard hit balls
after 100 pitches and the next start effect.

Aside from the compliment, this is an interesting research idea as well. I actually did reply in the article comments. I said:

An interesting question. I'll start looking at the various data sources. It will be
pretty easy to get stats per time-through-lineup, but I'm not sure about
time-through-lineup PLUS pitch count. In fact, I think I can get pitchers'
performance stats in games of 90+ pitches, but I don't know about their
performance ONLY on pitches after 90. Interesting idea.

Where’s that Costco receipt? Damn, I already lost it?! Oh, good—here’s my Farmer’s Market shopping list from March. Peppers ... squash ... tomatoes ... research piece on high-pitch-count hard-hit ball rates...

The last comment made still another interesting point:

With the cost of starting pitching injuries, (avoiding pitcher injuries) means
more money to assemble a good bullpen, and a good bullpen means
fewer injuries to your SPs, meaning ... more money to marshal
on position players.

I’m not sure if this is a virtuous circle or circular logic, or virtuously circular logic, but the premise that a good bullpen should result in fewer SP injuries was interesting to me. And, unlike high-pitch-count hard-hit ball rates, this one was easy to check.

I compiled all 30 teams’ bullpen stats for 2015-16, and scored the teams roto-style on three categories:

  • xFIP (an expected ERA metric allowing for league averages and park effects)
  • WPA/LI: The bullpens’ Win Probability Added, controlling for the Leverage Indexes of appearances (more credit given for tougher situations)
  • Shutdowns-Meltdowns: A counting stat adding relief appearances worth 0.06 WPA or more and subtracting appearances worth -0.06 WPA or less

After awarding points roto-style, I ranked the teams on their totals. Then I compared those overall rankings to the teams’ rankings in avoiding SP DL days lost, from fewest (CHW, 57) at the top to most (LA, 1,440) at the bottom.

The results: No connection. The top-10 bullpens included three teams that were top-10 in DL days, five teams in the middle-10, and two teams in the bottom-10. The middle-10 bullpens had four top-10 DL avoiders, two middle-10 and four bottom-10. And the worst bullpens had three top-10 DL avoiders, three middle-10 avoiders, and four bottom-10 avoiders.

The worst DL team, the Dodgers, was fifth-best among the bullpens. The best DL-avoiders, the White Sox, were third -worst in bullpens. A few teams matched in top- and bottom-10s: BAL and BOS in the top-10s of both, PHI, CIN and OAK bottom-10 in bullpens and injury avoidance.

This disconnect stands to reason. I think the logic of the idea works something like this:

  • Good bullpen creates manager confidence in his relievers
  • A manager confident in his relievers is likelier to hook starters sooner.
  • Starters hooked sooner have reduced workloads.
  • Starters with reduced workloads get fewer injuries.

The logic is valid, but one of the premises is not. The evidence I’ve read does not directly tie SP workloads—especially at the MLB level—to injuries. I’ve seen some good evidence that does tie injuries to high workloads for college and high-school-age pitchers. At MLB, I've seen pretty good evidence that injuries can be connected to throwing a lot of non-fastballs, and to a lot of high-stress workloads, like innings where the pitcher must throw a lot of pitches while battling through the threat of multiple baserunners. An early hook in those situations could help reduce injuries.

But for now, the evidence suggests good bullpens and SP injuries are unrelated.

Thanks to subscribers sempergumby, Reaper, andjohn and vjjr for your comments. Keep them coming! As I said, great ideas come from anywhere, and all the BHQ analysts really appreciate them.

Click here to subscribe

  For more information about the terms used in this article, see our Glossary Primer.